摘要
根据机械设备故障诊断本质特征和连续马尔可夫模型(CHMM)所具有的较强的时序模式分类能力的特点,提出了一种基于小波灰度矩向量与CHMM的滚动轴承故障诊断方法。从轴承振动信号提取一种量纲一的小波灰度矩向量作为特征参数,并训练几种故障状态的CHmm,再运用训练好的CHMM进行轴承的状态监测与故障模式的识别。诊断与对比实验表明该方法在故障样本少的情况下仍能进行准确训练与诊断。
According to the essential characteristics of fault diagnosis of mechanical equipment and traits that continuous hidden Markov model (CHMM) has the great ability of classifying time sequence pattern, a method of fault diagnosis based on the wavelet grey moment vector and CHMM for the rolling bearing was put forward. A nondimensional wavelet grey moment vector which was extrac- ted from the vibration signals was used as the feature parameters, and they were used to train the CHMMs of some fault states, then these trained CHMMs were applied to identify the states and fault patterns of bearings. At last, the experimental study verifies that the method of fault diagnosis for the bearings is effective with small training samples.
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2008年第15期1858-1862,共5页
China Mechanical Engineering
基金
国家自然科学基金资助项目(50675076
50545087)
国家重点基础研究发展计划资助项目(2005CB724101)
关键词
小波灰度矩向量
连续马尔可夫模型
模式识别
故障诊断
wavelet grey moment vector
continuous hidden Markov model (CHMM)
pattern identification
fault diagnosis